English Learners in California Schools:
Unequal
resources, unequal outcomes
Patricia Gándara
University of California, Davis
Russell Rumberger
University of California, Santa Barbara
Julie Maxwell-Jolly
University of California, Davis
Rebecca Callahan
University of California, Davis
Citation: Gándara,P.,
Rumberger, R., Maxwell-Jolly, J. and Callahan, R.,
(2003, October 7). English Learners in California Schools: Unequal
resources, unequal outcomes. Education Policy
Analysis Archives, 11(36).Retrieved [Date] from
http://epaa.asu.edu/epaa/v11n36/.
|
Abstract
The Williams vs the State of California class action suit
on behalf of poor children in that state argues that California
provides a fundamentally inequitable education to students based on
wealth and language status. This article, an earlier
version of which was prepared as background to that case, reviews
the conditions of schooling for English learners in the state with
the largest population of such students, totaling nearly 1.6
million in 2003, and comprising about 40 percent of nation’s
English learners. We argue, with evidence, that there are seven
aspects of the schooling of English language learners where
students receive an education that is demonstrably inferior to that
of English speakers. For example, these students are assigned to
less qualified teachers, are provided with inferior curriculum and
less time to cover it, are housed in inferior facilities where
they are often segregated from English speaking peers, and are
assessed by invalid instruments that provide little, if any,
information about their actual achievement. We end with suggestions
for ways in which teachers, administrators, and policymakers can
begin to address these inequities, even while legal remedies may
remain in the distant future.
|
Introduction
The fact that the United States remains an immigrant nation is
nowhere more apparent than in our public schools where an
increasing percentage of students are English learners. In 2000-01
these students represented ten percent of all students in the
United States, and 25 percent of California’s public school
population (Kindler, 2002). In the nation they numbered 4.6
million; while in California alone, they were more than 1.5 million
students. Most English learners both in the U. S. (79%) and in
California (83%) speak Spanish as their primary language. The
second largest language group in both California and the U.S. is
Vietnamese, however they account for only 2 percent and 2.5 percent
respectively. California is also by far home to more limited
English students than any other state. Thirty-three percent of all
of the nation’s English learner students live in that state;
the next largest concentration is in Texas, with 12 percent of the
total (Note 1). How best to
educate these students continues to be a highly controversial topic
and the source of considerable policy debate. However, with such a
large population of English learners, it is surprising how little
attention is actually paid to the basic learning resources these
student receive in California, and in the nation.
English learners are distributed throughout the schools, from
kindergarten to grade 12. One out of four students in the public
schools in California is an English learner, but one out of three
of the students in the elementary grades lacks proficiency in
English (Rumberger & Gándara, 2000, Table 1). There are
very few California schools that report having no English learners
among their student population (see Table 14). Today, the typical
California school is composed of both English learners and English
speakers, and in many schools more than one-quarter of the student
body is not fluent in English. Although most English learners are
found at the elementary school level, a larger proportion of
English learners (hereafter also referred to as ELs or EL students)
is found in secondary schools than commonly believed. More than 18
percent of California’s secondary school students (500,000
plus) are English learners (Rumberger & Gándara, 2000,
Table 1). Proportionately, the number of English learners in
secondary schools has been growing at a faster rate than the number
in elementary schools (California Department of Education, Language
Census 2001). The increase in the population of these secondary
level English learners presents a particular challenge for both the
students and the schools that serve them. This is principally
because older children have less time to acquire both English and
academic skills in order to get ready for high school graduation
and to prepare for post-secondary options. Unfortunately the unique
needs of these older EL students are often even more overlooked
than those of their younger peers. This article assesses the
condition of education for English learners in California, and, we
believe, has significant implications for the nation.
Organization of this Article
We begin this article with an examination of the achievement
data on English learners in California. We think it is first
important to establish the degree to which these students’
achievement represents a challenge to the overall productivity and
welfare of the state’s education system. In other words, we
attempt to make the case that the achievement gaps are so wide that
they threaten the well-being of the state and its economy, and
therefore should be a concern to everyone. We then follow with a
discussion of seven factors that we argue contribute significantly
to this situation. These seven factors are not exhaustive of the
problems faced by EL students, nor can they be neatly
compartmentalized. Some, like the shortage of skilled teachers,
represent both input shortcomings (e.g., insufficient
numbers of qualified teachers) as well as process problems
(e.g., inadequate instruction in the classroom) simultaneously.
Therefore, we present these factors roughly in the order in which
we think they affect the condition of schooling for English
learners and are amenable to policy intervention. We end with a set
of recommendations for addressing these issues. While this work
grew out of a major class action lawsuit in California, we
acknowledge that legal remedies almost certainly lie in the fairly
distant future, and more general social change perhaps beyond that.
Therefore our recommendations speak directly to the kinds of
actions that school personnel might undertake in the shorter
term.
Achievement of English learners
Data from a variety of sources reveal that the academic
achievement of English learners lags considerably behind the
achievement of English background students. We examined the
achievement of English learners using a number of different
measures and data sets –including data from the Early
Childhood Longitudinal Study (ECLS) of the U.S. Department of
Education, the American Institutes for Research Implementation of
Proposition 227 Study (Parrish, et al., 2001; 2002) and the
California Department of Education published data. (Note 2) At the same time that we present
analyses of existing data on student achievement for English
learners, we do so fully acknowledging the serious limitations of
achievement scores based on tests administered in English to
students who do not speak English well, or at all. We discuss below
the issues associated with such assessment.
Stanford 9 Achievement Scores
Between the years 1998 and 2002, the state used the Stanford
Achievement Test, Version 9 (SAT9)—a national, norm-
referenced, English-only achievement test—as the primary
means to assess the academic achievement of California’s
students. (Note 3) In spite of
the fact that we disagree with the state’s decision to use
this test for students who do not speak enough English to
understand it, we provide an analysis of the achievement of English
learners vis-à-vis their English-speaking peers as these same
test scores are routinely reported as accountability measures in
the state.
A persistent gap in test scores is a major factor in the school
experience of English learners. As a group they continue to perform
more poorly than English-speaking students throughout their entire
school career. This is clearly illustrated by the SAT 9 English
reading scores across grade levels (see Figure 1). As expected,
English learners who, by definition, are not yet proficient in
English, have low reading scores across all grade levels. Language
minority students who enter school already proficient in English
(Fluent English Proficient or FEP) start out comparable to native
English speakers, but by third grade they fall behind and never
catch up. Students who enter the schools as English learners and
who are subsequently reclassified as proficient (R-FEP), also start
out comparable, but by 5th grade they fall below native English
speakers, and by 7th grade they fall even further behind these
students. Such results challenge the belief that if English
learners simply demonstrated “proficiency” in English
–as defined by early scores on the SAT 9 test—the
achievement gap would disappear.
Even though this analysis shows a sizeable and growing
achievement gap between English speaking and non-English speaking
students across grade levels, there are some suggestions in the
data that the gap has narrowed slightly in recent years. To
investigate this issue, we examined SAT9 reading test scale scores
between the years 1998 and 2001 compiled by Parrish et al. (2002)
as part of the American Institutes for Research (AIR) year 2
evaluation of proposition 227. Scale scores show growth in
achievement over time based on a common metric. Thus it provides a
good indication of the amount of learning that has taken place over
time.
Figure 1. 2001 California SAT9 Reading Test
Scores by Grade Level and Language Background
SOURCE: California State Department
of Education, California Standardized Testing and Reporting
(STAR) Program. Retrieved February 7, 2002 from the World Wide
Web:
http://star.cde.ca.gov/star2001/default.htm
The AIR evaluation team had access to individual student test
scores for all the students in California for the years 1998
through 2001 by language classification. (Note 4) The evaluation team examined changes
in test scores between 1998 and 2001 for each grade level and for
three synthetic cohorts of students: (Note 5) (1) a cohort of students who were
enrolled in grade 2 in 1998, grade 3 in 1999, grade 4 in 2000, and
grade 5 in 2001; (2) a cohort of students who were enrolled in
grade 4 in 1998, grade 5 in 1999, and grade 6 in 2000, and grade 7
in 2001; and (3) a cohort of students who were enrolled in grade 8
in 1998, grade 9 in 1999, grade 10 in 2000, and grade 11 in 2001.
In order to compare non-overlapping cohorts, we replaced the second
cohort with one that began when students were enrolled in grade 5
in 1998. One of the innovations of Parrish and his colleagues is
that they compared English-only students with the combined group of
current English learners and former English learners who were
reclassified as Fluent English Proficient (R-FEP) in order to
better assess the progress of all students who first entered
California schools as English learners. Because an increasing
number of EL students become proficient in English as they progress
through school and are reclassified as fluent English speakers, the
number of EL students tends to decrease among older grade cohorts
while the number of R-FEP students tends to increase.
The results, illustrated in Figure 2, again show a sizeable
achievement gap between English only students and current/former
English learners. Both groups show more achievement growth in the
early years than in the later years, which reflects the increasing
difficulty of learning higher levels of more academic English
(Scarcella & Rumberger, 2000). The data show a slight
narrowing of the achievement gap across all three cohorts, as
Parrish, et al. note in their evaluation study (Parrish, et al.,
2002, page III-15). For example, the achievement level of English
only students improved from 581 points in grade 2 to 658 points in
grade 5, an increase of 77 points, while the achievement level of
English learners and former English learners improved 80 points. As
a result, the achievement gap narrowed by 3 points. Among all three
cohorts and three subjects (reading, language, and math), the 227
evaluation team found that the achievement gap narrowed by 1 to 8
points (Parrish, et al., 2002, Exhibits 10, 13, 16).
Figure 2. SAT 9 Reading Scores by Grade Cohort
and Language Classification, 1998-2001
SOURCE: Parrish, et al. (2002),
Exhibit 1.
It is interesting to note that the greatest achievement growth
for the grade 2 cohorts occurred in schools that offered bilingual
instruction before Proposition 227 or continued to offer bilingual
instruction after Proposition 227 (Figure 3). In addition, the
slight narrowing of the achievement gap between English only and EL
and former EL students noted above was due to reductions in the
achievement gap in those two types of schools, while in schools
that never offered bilingual education, there was no reduction in
the achievement gap.
Figure 3. Reading Achievement Gains for Grade
2-5 Cohort by Language Group and Instructional Model
NOTE: EO is English only; EL is
English learner; R-REP is reclassified Fluent English Proficient
(formerly EL).
SOURCE: Parrish, et al. (2002), Exhibits 10 and
19.
Despite these improvements, the achievement gap remains large
and increases at the higher grades. To illustrate, in grade 5, when
many students have completed elementary school, the left-most
horizontal line in Figure 2 shows that current and former English
learners are reading at the same level as English only students
between grades 3 and 4, a gap of about one and one half years. By
grade 8, when most students have completed middle school, the next
horizontal line shows that current and former English learners are
reading at the same level as English only students in grade 6, a
gap of about 2 years. By grade 11, the right-most horizontal line
shows that current and former English learners are reading at the
same level as English only students between grades 6 and 7, a gap
of about 4 and one half years.
California High School Exit Exam
The California High School Exit Exam (CAHSEE) is a major element
of California’s education accountability system. All students
in the class of 2004 and beyond were expected to pass the exam in
order to receive a high school diploma until very recently when the
State Board of Education authorized a delay in implementation of
sanctions until 2007. The data provide a hint as to why the class
of 2004 received this 11th hour reprieve. The exam is a standards-
based, criterion-referenced test that is designed to ensure that
all California high school graduates have a similar set of
fundamental skills in English language arts and mathematics
(California Education Code section 60850-60859). The need for
improving the education provided by California’s high schools
is undeniable. Although accountability measures may be necessary to
this effort, there is early evidence that the CAHSEE presents
exceptionally high stakes for EL students.
Although the test is a basic skills examination pegged to early
high school standards, (Note
6) by the end of their sophomore year, only 48 percent of
students from the class of 2004 had passed it. However, only 19
percent of English learners had passed at this same point
(California Department of Education, 2002, Attachment 1).
School Readiness
One reason for the underachievement of English learners is that
they begin school significantly behind their English-speaking
peers. Data from the Early Childhood Longitudinal Study (ECLS) show
that about half of California kindergartners from English speaking
backgrounds scored above the 50th percentile in fall assessments of
language, mathematics, and general knowledge. However, no more than
17 percent of kindergartners from non-English speaking backgrounds
scored above the 50th percentile (see Figure 4). One reason for
this disparity is that many English learners begin school without a
sufficient understanding of oral English that English background
students acquire naturally in their home environment. According to
the ECLS data, more than 60 percent of English learners who entered
California kindergartens in the fall of 1998 did not understand
English well enough to be assessed in English. And even after one
year of school, 38 percent of the students were still not
proficient enough in English to be assessed. (Note 7)
Figure 4. Cognitive Skills of California
Beginning Kindergartners by Language Background, Fall 1998
Note: Results are weighted
(C1CW0).
SOURCE: ECLS base year data for California public school
kindergarteners (N=2826).
Teachers and schools make judgments about students’
abilities based on the information available to them, including
test scores. Schools make class placement decisions based, at least
in part, on students’ standardized test scores. Moreover,
when the teacher does not speak the language of the child, cannot
communicate with the child’s family, and has little other
information to rely on, test scores can take on even greater
importance. Students who score low on tests are likely to be placed
in remedial education, even though such a placement is unlikely to
help students close the educational gap with their mainstream
peers. (Gottlieb, Alter, Gottlieb, & Wishner,1994; Skirtic,
1991). In Hobson v Hansen (269 F. Supp.401, 490; DDC 1967),
the Washington DC Superior Court noted in a major test case on the
viability of curriculum tracking as an educational practice that
“a sixth grade student nourished on a third-grade curriculum
is apt to finish the year with a third-grade education. .
.”.
Conditions of Inequity for English Learners
The achievement gap between English learners and their English-
only counterparts can be attributed, in part, to a number of
inequitable conditions that affect their opportunities to learn.
Our own research, combined with a review of the research of our
colleagues, leads us to identify seven primary areas in which these
students appear to receive a significantly inferior educational
experience, even when compared to other low-income students in the
public schools.
(1) Inequitable access to appropriately trained
teachers
English learners are more likely than any other children
to be taught by teachers with an emergency credential. There is
reason for concern about the low percentage of teachers who are
qualified to teach these students. An increasingly large body of
research has established that teachers with good professional
preparation make a difference in students' learning (Darling-
Hammond, 2002; Haycock, 1998; Sanders & Horn, 1995; Sanders
& Rivers, 1996). Moreover, a recent study conducted in Los
Angeles Unified School District (LAUSD) investigated the
relationship between English learner student achievement gains and
the credential held by the teachers who taught them in 29 schools
and 177 classrooms with large numbers of EL students. Hayes and
Salazar (2001) found that "state/district authorization of teachers
does have an impact on student outcome. For example, [Model B (Note 8)] students of teachers
holding no state or district authorization achieved largely
negative or very small positive. . . . adjusted gains in reading
and language" (pp. 37-38). (See Table 1). A follow up study of
grades 1 – 3 classrooms in the same schools during the
subsequent school year (2001) found again that “students of
credentialed teachers out-performed students of emergency permitted
teachers” (Hayes, Salazar & Vukovic, 2002, p. 90).
Table 1. Actual and Adjusted Gains by Teacher Authorization
Grade 2, Selected Schools, LAUSD
| |
Reading |
Language |
| |
Actual Gains |
Adjusted Gains |
Actual Gains |
Adjusted Gains |
| BCLAD |
1.8 (n=142) |
1.6 (n=142) |
4.1 (n=148) |
2.4 (n=148) |
| CLAD/LDS |
2.0 (n=32) |
2.7 (n=32) |
1.0 (n=34) |
0.4 (n=34) |
| SB1969 |
* |
* |
* |
* |
| A Level@ |
1.8 (n=155) |
1.6 (n=155) |
0.3 (n=155) |
-1.5 (n=155) |
| No Authorization |
-2.4 (n=74) |
-2.9 (n=74) |
0.5 (n-93) |
-1.8 (n=93) |
*Actual and adjusted gains were not reported here
due to the small sample size.
@ LAUSD certifies language competencies of its teachers if they do
not already hold a BCLAD; A Level indicates fluent bilingual.
Source: Hayes & Salazar (2001), page 36
Whereas 14 percent of teachers statewide were not fully
credentialed, 25 percent of teachers of ELs were not fully certified
(Rumberger, 2002). Figure 5 shows that as the concentration of ELs
in a California school increases, so too does the percentage of
teachers holding emergency credentials. Inasmuch as Figure 5 holds
poverty constant, we would expect to see a flat line if the
discrepancy in credentialed teachers were purely a function of
poverty. These data show that English learners are significantly
less likely to have a fully credentialed teacher than other low-
income non-EL students. We will demonstrate that this is largely a
problem of uneven distribution of qualified teachers among
California's schools and classrooms.
Figure 5. The Relationship between the Percent
of English Learners and the Percent of Teachers with Emergency
Credentials, Holding Constant the Percent of Students on Free or
Reduced Lunch, California Schools, 1999-2000
Note: Relationship estimated from the
regression equation: 3.553 + .119*LUNCH + .095*ELL (N=6039), with
LUNCH = 48.6 (sample mean).
Source: 1999-2000 API Growth Data File. Retrieved October 4, 2000
from WWW: http://api.cde.ca.gov/datafiles.html.
Authorizations to Teach English learners
The current state of the art of teaching EL students employs
three central methodologies for English learner instruction. The
first strategy, specially designed academic instruction in English
(SDAIE), is defined as “a set of systematic instructional
strategies designed to make grade-level and advanced academic
curriculum comprehensible to English learners with intermediate
English language proficiency” (California Commission on
Teacher Credentialing, 2001, p. 26). Another means of teaching EL
students is through their primary language. Although the principal
goal is to provide access to the core curriculum, in reality, this
involves a continuum of strategies, from using the student’s
primary language solely for clarification of concepts presented in
English to actually providing academic instruction in the primary
language. A third strategy is English language development (ELD).
It is “systematic” instruction of English language that
is designed to (1) promote the acquisition of English-listening,
speaking and reading and writing skills—by students whose
primary language is other than English, and (2) provide English
language skills at a level that will enable equitable access to the
core curriculum for English learners once they are presented with
academic content. (CTC, 2001, p. 3).
The California Commission on Teacher Credentialing (CTC) issues
basically two EL credentials meant to ensure that teachers have
skills in some or all of the above instructional strategies (see
Table 4): the Bilingual, culture, language and development
credential (BCLAD) and the Culture, language and development
credential (CLAD). Many California teachers of English learners
hold earlier versions of these specialized credentials that are
generally considered equivalent and authorize them to teach English
learners (Note 9).
Skills, knowledge, and instructional settings approved for each
authorization
The most rigorous of the credentials, the bilingual, culture,
language, and development (BCLAD) certification, requires that
teachers have expertise in the areas of: 1) language structure, 2)
methodology for first and second language development, and 3)
cross-cultural competency. BCLAD teachers must also demonstrate
competency in three additional spheres: 4) methodology for primary
language instruction, 5 & 6) knowledge of a particular culture
and language of emphasis. Many BCLAD teachers earn their expertise
through a Master’s Degree program or through a credential
program with an emphasis on teaching English learners infused
throughout the program’s coursework and field placements.
BCLAD authorization requires extra expertise because it
authorizes teaching in all settings with English learners. These
teachers are authorized in the various methods of EL instruction
for conveying academic content and promoting English language
proficiency including primary language methods, "specially designed
academic instruction in English" (SDAIE), and English language
development (ELD). As speakers of a second language, these
teachers tend to be more sensitive to the issues surrounding the
acquisition of a second language and can communicate with students
and parents in at least one language other than English. Thus,
these teachers have a variety of skills to address a range of EL
students’ educational needs.
The next most comprehensive authorization, the CLAD certificate
or credential includes the first three skill areas required of the
BCLAD teacher: 1) language structure, 2) methodology for first and
second language development, and 3) cross-cultural competency.
Expertise in these areas is gained through a set of four college
courses—or by passing exams on this content. CLAD teachers
should have some experience of learning a second language but are
not required to have a command of that language or culture that is
required for BCLAD certification. CLAD holders are authorized to
teach subject matter to EL students using SDAIE and other English
language methods, and to teach English language development.
Staffing EL classrooms with BCLAD or CLAD teachers allows
English learners to remain in self-contained classrooms. Classrooms
without CLAD or BCLAD teachers may require that EL students be
removed for ELD (or academic support), so called pull-out
instruction (Brisk, 1998). Despite being ubiquitous in English
learner education, pull out instruction has been found to be among
the least successful of instructional strategies for these students
(Lucas, 1997; Ovando & Collier, 1998). Although BCLAD
certification is the most comprehensive, it is also the rarest:
Only 5% of California teachers who instruct English learners have a
full credential with BCLAD authorization
(UC LMRI, 2003).
“Teacher in training” Status
According to the California Commission on Teacher Credentialing,
the most widely used option to teach English learners is the
“teachers in training” status, which does not require
any certification. Rather, teachers in training are permitted by
the California Department of Education to teach English learners
using ELD and SDAIE methods based upon a mere agreement to obtain
the requisite training for certification within two or three years.
Teachers in EL classrooms who sign agreements that they are
participating in or will obtain the requisite training are
conditionally allowed to continue in their positions by the CDE.
Unlike the various other certifications offered, the teachers in
training status is not monitored by the Commission on Teacher
Credentialing. Rather, this status was developed by the California
Department of Education as part of a “plan to remedy”
the shortage of teachers certified to teach ELs in school districts
that were found by the Coordinated Compliance Review (CCR) to be
out of compliance with matters concerning English learners.
However, instead of remedying the shortage of certified teachers,
the CDE’s re-labeling of untrained teachers has largely
reinforced the status quo. Thus far, CDE monitoring and enforcement
of these agreements has not resulted in any substantial reduction
of the numbers of “teachers in training”.
Supply of EL Authorized Teachers in California
To determine whether the problem is one of sufficient numbers of
teachers qualified to teach English learners, or simply a
maldistribution of qualified teachers, we analyzed figures from the
California Basic Educational Data System (CBEDS) for the year 1999-
2000. CBEDS conducts an annual survey of every professional
educator working in the public school system. Teachers are asked to
indicate the type of California teaching credential they hold,
including whether it is a "full" credential or an "emergency"
credential. Teachers are also asked to indicate all the areas that
their credential authorizes them to teach. We identified all
teachers who indicated that they were authorized to teach in
bilingual, English language development, or specially designed
academic instruction in English (SDAIE) classes. We then compared
the number of teachers with such special authorization to teach ELs
with the number of EL students, both in the state as a whole and in
each school that enrolled English learners. We also compared these
figures with data on students who were not English learners and
teachers without authorization to teach English learners.
Table 2. California Students and Teachers by Language
Background, 1999-2000
| |
English Learner |
Other |
Total |
| Students |
1,480,406 |
4,471,206 |
5,951,612 |
| |
|
|
|
| Teachers, including emergency
permits/waivers |
79,215a |
212,840 |
292,055 |
| Students per teacher |
18.7 |
21.0 |
20.4 |
| Teachers per 100 students |
5.4 |
4.8 |
4.9 |
| |
|
|
|
| Teachers excluding emergency
permits/waivers |
75,687 a |
175,781 |
251,468 |
| Students per teacher |
19.6 |
25.4 |
23.7 |
| Teachers per 100 students |
5.1 |
3.9 |
4.2 |
| |
|
|
|
| Fully credentialed bilingual/ELD teachers |
69,305 b |
|
|
| Student per teacher |
21.4 |
|
|
| Teachers per 100 students |
4.7 |
|
|
| |
|
|
|
| Fully credentialed bilingual teachers |
26,539 c |
|
|
| Students per teacher |
55.8 |
|
|
| Teachers per 100 students |
1.8 |
|
|
- Teachers authorized in any way to teach bilingual education,
English Language Development, or specially designed academic
instruction in English (SDAIE), including those with SB1969
authorizations.
- Teachers authorized to teach bilingual education or English
Language Development.
- Teachers authorized to teach bilingual education.
SOURCE: 1999 CBEDS and 2000 Language Census.
The figures in Table 2 show that in the state as a whole, there
were almost 6 million students and almost 300,000 teachers, which
represents 20 students per teacher or 5 teachers per 100 students.
There were also about 1.5 million English learners and about 79,000
"EL" teachers, that is, those with some kind of special
authorization (BCLAD, CLAD, 1969/395) to teach them through the
primary language and/or ELD, and/or SDAIE. Ignoring for the moment
whether all of these authorizations are adequate to the task of
teaching English learners, this represents about 19 students per EL
teacher or more than five EL teachers per 100 EL students. These
figures suggest that there are slightly more teachers with some
specialized preparation per EL student in the state than the
statewide student/teacher ratio. The same conclusion can be drawn
if a similar analysis is done with only teachers who are fully
authorized to teach English learners: there are actually more fully
authorized EL teachers in the state per EL student than there are
fully credentialed (non-EL) teachers per non-EL student. However,
the language census data indicating how many EL students are
actually taught by authorized CLAD or BCLAD teachers paints a
somewhat different picture. These data indicate a statewide average
of only 4.2 CTC authorized EL teachers per 100 English language
learners (California Department of Education, Education
Demographics Office, Spring 1999 Language Census).
Comparing the numbers of teachers with the most rigorous
training to teach English learners, those with BCLAD, bilingual
specialist, or BCC credentials, the picture changes dramatically.
Based on the same procedure as above, there are only 1.9 fully
credentialed BCLAD equivalent teachers (i.e., those with the
most comprehensive credential) for every 100 EL students versus 3.8
fully credentialed teachers per 100 non-EL students, or half as
many. Under this scenario, the state would need another 26,000
teachers with the most comprehensive credentials to reach the same
proportion as for non-EL students taught by teachers with the most
comprehensive training.
Distribution of EL Teachers in California
While this statewide picture suggests that there are sufficient
numbers of EL teachers with at least some authorization to teach
English learners, it does not indicate how those teachers are
distributed among schools. To investigate this issue, we classified
schools based on the number of fully credentialed EL teachers they
had for every 100 EL students. We divided schools into four groups:
(1) schools with no EL teachers, (2) schools with a ratio of fewer
than 2.5 fully credentialed EL teachers per 100 EL students--half
the state average, (3) schools with a ratio between 2.5 and 7.5,
and (4) schools with a ratio of more than 7.5--50 percent above the
state average. We then computed how many schools were in each
category and how many EL students attended those schools (Table
3).
Table 3. Number of English Learners by EL Teacher/Student
Categories and Level, 1999-2000
| Fully Authorized EL teachers per 100
EL students |
Elementary |
Middle |
High |
| Number |
Percent |
Number |
Percent |
Number |
Percent |
| No EL teachers |
18,689 |
1.9 |
5,703 |
2.4 |
1,675 |
0.7 |
| Fewer than 2.5 |
193,205 |
19.7 |
81,954 |
35.3 |
74,119 |
31.3 |
| 2.5 to 7.5 |
610,629 |
62.3 |
120,153 |
51.7 |
132,402 |
55.8 |
| Greater than 7.5 |
157,331 |
16.1 |
24,671 |
10.6 |
28,933 |
12.2 |
| Total |
979,854 |
100.0 |
232,481 |
100.0 |
237,129 |
100.0 |
Source: 1999 CBEDS and 2000 Language Census
At the elementary level, more than 200,000 English learners--20
percent of the total--attend schools with 2.5 or fewer EL teachers
per 100 English language learners. At the middle school level, more
than 85,000 ELs attend such schools-- almost 38 percent of the
total. At the high school level, more than 75,000 attend schools
with such low numbers of qualified EL teachers--almost one-third
of all high-school EL students. Counting English learners who
attend other types of schools (e.g., alternative, continuation,
etc.), more than 390,000 English learners in California--one out
of every four--attends a school with fewer than half the state
average of teachers with specialized authorizations to teach
them.
Another indication of the shortage of teachers with the
appropriate training to teach English learners is revealed from an
analysis of the 2000 Class Size Reduction (CSR) teacher survey
(Stecher & Bohrnstedt, 2002). According to those data, 37
percent of all teachers who taught grades 1-4 in 2000 held a CLAD
credential, 10 percent held a BCLAD credential, and 45 percent held
either a CLAD or BCLAD (see Table 4). In general, the higher the
concentration of English learners in the classroom, the higher the
proportion of teachers who held at least some authorization to
teach them. Yet among classrooms where a majority of students are
English learners, only about half of the teachers held an
appropriate EL credential. Using data on the proportion of English
learners in each type of classroom, we estimate that only 53
percent of all English learners enrolled in grades 1-4 in
California in the 1999-2000 school year were taught by a teacher
with any specialized training to teach them (Note 11). If we assume that
teachers with BCLAD credentials have the most appropriate training,
only 22 percent of all English learners enrolled in grades 1-4 had
such a teacher in 2000.
Table 4. Percent of Teachers in Grades 1-4 with CLAD and
BCLAD Credentials By Classroom Concentration of English Learners,
2000
| Percent English Learners in the classroom |
Percent of all English Learners |
CLAD |
BCLAD |
CLAD or BCLAD |
| 0 |
0 |
25 |
2 |
27 |
| 1-25 |
17 |
47 |
3 |
47 |
| 26-50 |
20 |
46 |
11 |
54 |
| 51-100 |
63 |
29 |
30 |
54 |
| Total |
100 |
37 |
10 |
45 |
NOTE: Results are weighted.
Source: 2000 Class Size Reduction Teacher Survey (N=774).
Class size reduction had some largely unanticipated consequences
for EL students because of the relative concentration of English
learners in the state's poorest schools. The migration of
credentialed teachers away from these schools to those in more
affluent areas with better working conditions was a significant
feature of the class size reduction initiative in California
(Stecher & Bohrnstedt, 2002). For example, the percentage of
teachers not fully credentialed in schools with the smallest
proportion of English learners (less than 8 percent) only increased
from .3 percent in 1995-96 to 4.0 percent in 2000-01 (see Figure
6). However, the percentage in schools with the greatest proportion
of English learners (40 percent or more) increased from 3.7 percent
to 23.9 percent over the same five-year period. As a result,
schools with the most English learners benefited the least from
class-size reduction, at least in terms of access to fully
credentialed teachers.
Figure 6. Percentage of Public K-3 Teachers
Not Fully Credentialed by School Quartiles of English Learners:
1995-96 to 2000-01
SOURCE: Brian M. Stecher and George
W. Bohrnstedt (Eds.), Class Size Reduction in California:
Findings from 1999-00 and 2000-01 (Palo Alto: American
Institutes for Research, 2002), Table B.17.
At the same time that EL students are less likely than others to
have a qualified teacher, the challenges associated with teaching
them are even greater than for the typical student. The large
number of English learners who are immigrants frequently come from
circumstances in which their early lives and education have been
disrupted by war, loss or estrangement of family members, poverty,
and residential mobility (Ruiz de Velasco & Fix, 2000; Olsen,
1998). As such, teachers must know how to intervene educationally
with students whose personal and educational backgrounds are
significantly different from the mainstream English-speaking
student. Moreover, the age and grade placements of these students
in U.S. schools often do not match their skill levels because of
varying educational experiences in their countries of origin (Ruiz-
de-Velasco & Fix, 2000).
Necessary Competencies of EL Teachers
Wong Fillmore and Snow’s (2000) study, entitled What
Teachers Need to Know about Language, outlines the critical
knowledge base that teachers must have for language learning in
order to effectively teach children who do not speak standard
English. They note that teachers need to know the units of language
and how they operate differently across languages and dialects. For
example, knowing how tense and plurality are formed in the
child’s native language can help the teacher to uncover
difficulties in English and facilitate learning for ELs. Wong
Fillmore and Snow also argue that by knowing the fundamental
characteristics of words in the primary language of the student,
the teacher can facilitate more rapid acquisition of English
vocabulary and word construction. They point out, for example, that
if a teacher can explain that the suffix idad in Spanish has
the same consistent meaning as ity in English, the
student’s vocabulary and word usage can be expanded
significantly. These authors also assert that teachers must
understand the norms for language usage in the primary culture of
the student in order to know how to encourage English learners in
their acquisition of English. Another critical competency that Wong
Fillmore and Snow argue teachers must have is a clear understanding
of what constitutes academic English and how to support the
acquisition of this particular form of the language for English
learners. Academic English is the language of texts and often of
tests, and it is not normally acquired in the course of
conversation outside of academic contexts. For students who are not
likely to “absorb” this form of English discourse in
their homes or communities, it must be explicitly taught.
Of course, one of the most controversial of all topics in
education is the best method for teaching reading. Many experts
argue that there is no single best method. Rather there are a
number of strategies that are more or less effective with different
students at different points in the process of learning (Snow,
Burns & Griffin, 1998), and it requires the expertise of a
well-trained teacher to know which strategy to use when, and with
which children. There is even less agreement, however, on how best
method to teach English learners to read in a language they do not
understand–English. The National Research Council (Snow,
Burns & Griffin, 1998) concluded that if reading instruction is
not done initially in the primary language of the child, then
educators should consider delaying it until English is acquired.
This points out the degree to which the field continues to depend
upon the skills of highly qualified teachers to make judgments
about how best to teach reading to English learners. There simply
are no “tried and true” strategies for teaching
children to learn to read in a language they do not understand, and
it is a vexing problem even to the experts in the field
Finally, there are significant issues associated with the
cultural backgrounds of immigrant and non-English speaking students
that bear on how they learn. Wong Fillmore and Snow (2000) argue
that a critical role for all teachers is to socialize students to
the demands of schooling. Certainly, this is made more difficult if
the teacher does not understand both the cultural and linguistic
norms of the students he or she is teaching. They point out that in
correcting some students, or encouraging others to participate in
linguistically-based activities in the classroom, teachers may
inadvertently squelch the motivation of English learners to
participate at all. Without understanding the cultural and
linguistic expectations of these students’ communities,
teachers can undermine their students’ learning by failing to
acknowledge culturally appropriate behavior. For example, many
teachers reward students for questioning behaviors and active
participation in discussion, but some immigrant students are
socialized to believe that such behavior is inappropriate in the
classroom (Olsen, 1997).
Teachers themselves have cited their need for greater expertise
in working with EL students. In a survey of all 1999-2000 graduates
of teacher credential programs in the California State University
system (total of 10,512) one fourth responded that they felt they
were only "somewhat prepared" or "not at all prepared" to teach
English learners (Office of the Chancellor, 2002). We note that
these are the "cream of the crop" of teachers of English
learners—those who have completed a full credential and in
most cases have training at least at the level of the CLAD
(Culture, Language and Development preparation (CLAD) credential (Note 12).
Many teachers of English learners also report difficulty in
communicating with the parents of their students. In a recent
Harris survey (Note 13), 23
percent of teachers of English learners reported that they had a
hard time communicating with their English learners’ parents
about their children’s educational progress and needs (Table
5). Not surprisingly, teachers with no special preparation to teach
English learners were more likely to report difficulty, while most
teachers with BCLAD credentials reported that they were able to
communicate with their students’ parents. It is also notable
that in the Hayes et al. (2002) study of the LAUSD implementation
of Proposition 227, the largest concern noted by non-English
speaking parents was lack of communication with teachers.
Table 5. Percent of teachers with EL students in their
classes who reported difficulty in communicating with parents by
teacher credential, January 2002
| Teacher credential |
Percent reporting difficulty |
| CLAD (or equivalent) |
25 |
| BCLAD (or equivalent) |
7 |
| SB-1969/395 Certificate |
10 |
| None |
30 |
| Total |
23 |
NOTE: Results exclude respondents who did not
answer question or answered “not sure.” Results are
weighted. Overall differences are statistically significant at .05
level or better.
Source: Harris Survey of a Cross-Section of California School
Teachers, January 2002 (N=820).
A common critique of teacher preparation programs, both in
California and elsewhere, is that the extant knowledge of how to
teach English learners is not often incorporated into teacher
preparation efforts (Wong-Fillmore & Snow, 2000; Reagan, 1997;
Tomas Rivera Center, 1994; Milk, 1990). In effect, we know
considerably more about how to prepare teachers than we act on in
schools of education. This is generally viewed as a problem in
translating research into practice. The reasons for this have been
debated at great length in the education literature (cf. Cochran-
Smith & Fries, 2001), but the only firm conclusion that can be
drawn is that there is a clear disjunction between research and
practice in teacher education. It is nowhere more painfully evident
than in the preparation of teachers for English learners where it
is commonly argued that the field lacks research-based methods,
when in fact this is not the case (cf. August & Hakuta, 1997;
Gándara & Maxwell-Jolly, 2000).
Working Conditions for EL Teachers
Given the opportunity, teachers vote with their feet for where
they want to work, and school conditions appear to influence this
vote. Recent research suggests that working conditions influence
teachers’ decisions about where to teach more than do
salaries (Hanushek, Kain, & Rivkin, 2001; Loeb & Page,
2000). Data for California demonstrate this clearly. Table 6
demonstrates that the differences between conditions in schools
with high and low concentrations of EL students are dramatic, even
with respect to characteristics that would not intuitively seem to
be related to the concentration of English learners. However, it is
evident that when working and learning conditions are poor, they
affect the attitudes of staff, and no doubt the ability of the
school to attract competent and amiable people to work there.
Table 6. Characteristics of the Environment of California
Elementary Schools by EL Concentration, Spring 1999
| |
25% or less |
Over 25% |
Total |
| Problems in neighborhood where school is
located: |
|
|
|
| Selling or using drugs or excessive
drinking in public |
16 |
50 |
29 |
| Gangs |
32 |
77 |
50 |
| Crime |
24 |
77 |
45 |
| School climate: |
|
|
|
| Parents not active in programs |
4 |
40 |
18 |
| Problem with overcrowding |
25 |
50 |
34 |
Note: Results are weighted (S2SAQW0). All column
differences are statistically significant at .05 level or
better.
SOURCE: ECLS base year data for California public elementary
schools (N=69).
Given the difficult working conditions and the added demands
placed on teachers of English learners, it would be expected that
both training and guidance on how to address these challenges would
be provided. The data, however, show otherwise. Teachers of English
learners are largely left to fend for themselves with inadequate
guidance, resources, and training.
(2) Inadequate professional development opportunities to help
teachers address the instructional needs of English
learners.
Although there are no studies that are able to demonstrate a
causal relationship between student achievement outcomes and
teacher professional development, the relationship between teacher
preparation and student achievement (see, for example, Darling-
Hamond, 2002) suggests that such a relationship probably does
exist. Moreover, a number of studies have demonstrated that good
professional development increases teachers’ sense of
competence and provides them with tangible strategies for better
meeting the needs of their students (Herman & Aguirre-Munoz,
2003; Herman, Goldschmidt, & Swigert, 2003). Given these
findings, teacher professional development has been a cornerstone
of many states’ education reform plans, including
California’s. Yet, surprisingly little emphasis has been
placed on the specialized needs of teachers of English
learners.
The instructional demands placed on teachers of English learners
are intense. They must provide instruction in English language
development while simultaneously or sequentially attempting to
ensure access to the core curriculum. Yet data collected for
California’s Class Size Reduction Study (Stecher &
Bohrnstedt, 2000) show that even where teachers are teaching a
majority of English learners, the professional development they
receive that is dedicated to helping them instruct these students
is minimal. The percent of professional development time that
teachers reported focusing on the instruction of English learners
in 1999-2000 ranged from 3 to 10 percent with a mean of only 7
percent (Table 7).
Table 7. Professional Development of Teachers in Grades 1-4
by Classroom Concentration of English Learners, 1999-2000
| Percent English Learners in the classroom |
Mean number of days |
Mean number of hoursa |
Percent of hours on teaching English
Learners |
| 0 |
3.5 |
28 |
3 |
| 1-25 |
3.6 |
30 |
8 |
| 26-50 |
3.3 |
32 |
9 |
| 51-100 |
3.8 |
35 |
10 |
| Total |
3.6 |
31 |
7 |
Number of hours estimated by recoding responses
(8 hours or less = 4 hours; more than 8 = 12 hours).
Note: Results are weighted.
Source: 2000 Class Size Reduction Teacher Survey (N=774).
These data are corroborated by several other recent studies.
Hayes and Salazar (2001), in their study of 177 classrooms in the
LAUSD, noted that teachers discussed “the problematic lack of
resources and training to assist them to provide quality services
to ELLs (p23).” A report on the results of a California
Department of Education (CDE, 1999) survey of every California
school district during the first year of Proposition 227
implementation showed that professional development to help
teachers with English learner instruction was one of the most
significant unmet needs in the aftermath of the passage of the
proposition. The later, more ambitious, study of the implementation
of Proposition 227 being conducted by American Institutes for
Research (Parrish, et al., 2001, 2002) likewise reports a similar
theme emerging from their investigation. The study documents a
significant lack of guidance from the state about the nature of the
instruction that should occur in the Structured English Immersion
classrooms, and as a result, “teachers were not provided
appropriate materials or guidance on how to use materials
appropriately” (Parrish, et al., 2001, p. 36). Again, in the
most recent report of this five-year study, researchers concluded
that, “Barriers to the implementation of the Proposition
include insufficient guidance for implementing regulations in the
law; confusion over what the law requires and allows; and lack of
clear operational definition for the various instructional
approaches for EL students. In particular, educators lack clarity
on what constitutes best practice within structured English
immersion instruction” (Parrish, et al., 2002, p. ix).
The University of California has provided professional
development for the state’s teachers through its Professional
Development Institutes (CPDIs). This is not the only professional
development activity in the state, in fact, many districts sponsor
extensive professional development programs, but it is the largest
state-wide effort, with more than 45,000 teachers participating in
these workshops in 2000-01. In that same year, a total of
$50,866,000 was provided for this purpose. Of this amount, only
$8,358,104 was earmarked for professional development in the area
of English Language Development (Office of the President,
University of California, 2002). This constituted about 16% of the
professional development budget, although English learners
constitute fully 25% of the students in the state, and as we have
argued, are the most educationally deprived by their schools of all
students. The AIR study found that only 18% of the teachers in
their sample had even heard of the ELD CPDIs, and only 8% had
attended one or more (Parrish, et al, 2002, p. IV-40), suggesting
that relatively little is done to disseminate information about
resources that may be available to teachers of English
learners.
3. Inequitable access to appropriate assessment to measure EL
achievement, gauge their learning needs, and hold the system
accountable for their progress (Note14)
While English learners must be incorporated into a state
accountability system in order to insure that their educational
needs are being met, the current system is of little value for
monitoring their academic progress.
English Language Testing of the Academic Progress of English
Learners
According to the National Academy of Sciences, “when
students are not proficient in the language of the assessment
(English), their scores on a test in English will not accurately
reflect their knowledge of the subject being assessed” (NRC,
1999, p. 214). Therefore such assessments provide neither accurate
data for accountability purposes, nor do they help teachers to
enhance their instruction. As the National Research Council noted,
“ if a student is not proficient in the language of the test,
her performance is likely to be affected by construct-irrelevant
variance –that is, her test score is likely to underestimate
her knowledge of the subject being tested” (NRC, 1999, p.
225). These tests can, moreover, have serious negative effects on
the schooling of English learners in at least two ways: (1)
positive changes in test scores over time can give the inaccurate
impression that students have gained subject matter knowledge when,
in fact, they may have simply gained proficiency in English. This
misperception that EL students are making academic progress can
lead schools to continue providing a curriculum that fails to
emphasize comprehensible subject matter. (2) On the other hand,
consistently low scores on tests can lead educators to believe that
students need low-level or remedial education, when in fact, they
may have mastered the curriculum in another language, but are
unable to express these competencies through an English language
test.
The current state accountability practice for English learners
is as follows:
- All EL students in Grades 2-11 must take the Stanford 9 (SAT 9)
(Note 15), a nationally
norm-referenced test in reading and math (and science and social
studies in the higher grades) administered in English, unless
parents or a guardian provides a written request for a waiver.
- Only “ELLs who have been in the district for 12 months or
more may not use nonstandard accommodations unless they have
individualized education plans (IEPs) or Section 504 plans that
include accommodations.” (Note 16)
- English learners who have been in a district for less than one
year (except for entering ninth graders in high school districts as
of 2000) are excluded from the Academic Performance Index (API). (Note 17)
- The API is used to measure each school’s performance
based on student test scores. Through the Governor’s
Performance Award (GPA) Program, there are monetary and non-
monetary awards based on positive changes in the API. In addition,
through the Immediate Intervention/Underperforming Schools Program
(II/USP), failing schools are identified for local and state
intervention to improve programs
- “Spanish-speaking English language learners who had been
enrolled in California public schools less than 12 months when
testing began [are] required to take the SABE/2 in addition to
taking the Stanford 9…” (Note 18)
- Finally, the CAHSEE is the gatekeeper of graduation that all
students, including English learners and other students with
exceptional needs, must pass in order to receive a high school
diploma (as of 2007).
The exclusive reliance on an English-language norm-referenced
achievement test for EL students is inappropriate for these
students (Note 19) and
violates several standards established by the authoritative
AERA/APA/NCME Standards for Educational and Psychological Testing.
(Note 20)
Research on second language acquisition shows that it takes
English learners on average between four to seven years to meet
various standards of English proficiency (Hakuta, Butler &
Witt, 2000).The burden is on the state to demonstrate that test
scores for English learners who have been in the United States for
less than four years are valid, yet the state has not made any
attempt to obtain information to shed light on this question (Note 21).
The only cautionary statement by the CDE on the interpretation
of standardized test scores appears on a web page and says:
“Since the Stanford 9 norming sample was representative of
the United States as a whole, it does not necessarily match
California's student population.” (Note22) There is no acknowledgement such as
that of the San Diego Education Association that explicitly states
that the California population is vastly different from the norming
group: “The norming sample, while representative of the
nation, does not reflect the huge diversity of California's student
population. For example, 40.5% of California's students are
Hispanic, but only 9.6% are in the Stanford 9 norming sample. While
24.6% of the state's students are of limited English proficiency,
only 1.8% are in the sample.” (Note 23) Since the test scores are reported
with respect to the national percentile rank (NPR), failure to
issue an explicit warning with respect to Hispanics and to English
learners is a clear violation of this standard.
The state requires the collection of data using a comparable
test in Spanish, the SABE/2. However, the state gives no guidance
on how this information might be used to augment information from
SAT 9. Indeed, while requiring SABE/2 for Spanish speaking students
who have been in the U.S. for less than 12 months, the Department
of Education explicitly rejects the use of SABE/2 in its Academic
Performance Index (API) and does not monitor its administration. (Note 24) Among the reasons for
this according to the Superintendent’s Advisory Committee for
the Public Schools Accountability Act of 1999 is that “SABE/2
is not aligned with state content standards”, leading to
“consensus in the API subcommittee to not include the SABE/2
in the 1999-2000 API.” (Note 25) That SABE/2 is not aligned with the
state content standards is uncontestable. But neither was the SAT 9
at the time the API system was developed. More recently, the SAT 9
has been augmented with new and revised items to bring it into
alignment with state standards, and in fact, a new test is being
developed that will be in even greater alignment. The same could be
done with the SABE/2 or another similar achievement test in
Spanish. Given the importance of API as a policy instrument in
driving rewards and sanctions for school academic performance, it
is indefensible that the state makes no provisions for the use of
information from a native language test, indeed one that the state
requires (on paper only) students to take. This point is made
particularly salient by the discrepancies between English learner
scores on the SABE/2 and on the SAT 9. Well over half (59%) of all
4th graders taking the SABE/2 reading test in 2001 scored at or
above the 50th percentile on this test, which was normed on a
Spanish-speaking population. In contrast, only 15 percent of
English learners in the 4th grade performed as well on the SAT 9 (Note 26). The tests are not
strictly comparable, but the discrepancies raise serious questions
about the appropriateness of current practice of educational
planning based on clearly faulty and incomplete information about
what EL students know and need to know.
California English Language Development Test (CELDT)
The CELDT test of English language development was developed to
make it possible for educators to assess the level of English
proficiency of their English learners. Administered for the first
time in 2001, the test, developed by CTB-McGraw Hill, and designed
to assess English learners in grades 1 through 12 on reading,
writing, comprehension, and oral English along a continuum of five
levels, from Beginning (1) to Advanced (5). The purpose of the test
is to monitor the growth of students’ English skills over
time and to provide a single statewide measure useable for both
program placement and reclassification to the status of English
proficient. Prior to the development of the CELDT, school districts
used a variety of different tests which did not align with each
other and provided no consistent basis for monitoring student
progress (Gándara & Merino, 1994). While the CELDT was
warmly received by educators of English learners as a potentially
useful assessment tool, it suffers from at least two major
limitations. It was so time-consuming to administer that it placed
a significant burden on schools. In addition, scores were not
available in a timely enough fashion to allow educators to use the
information for program purposes (Note 27).
As a result of
these limitations, the CELDT has been revised by CTB-McGraw Hill.
The restructured CELDT, form C, is being administered at the time
of this writing. To what extent this makes it impossible to compare
scores from 2001 to 2003 is not yet known.
California High School Exit Exam
EL students are much less likely to pass the High School Exit
Exam (CAHSEE) than are English speakers (Note 28). Students with exceptional
needs as defined in Title 20 of federal law may take the exams with
accommodations to meet their special needs. However, English
learners do not have exceptional needs according to this definition
and do not therefore qualify for accommodations in the state of
California. The law does allow for districts to defer the
requirement that students pass the exam until the pupil has
completed six months of instruction in reading, writing, and
comprehension in the English language. Nonetheless, no student,
including those who are still classified as English learners, will
receive a high school diploma without passing the exit examination
in English.
An important feature of the law authorizing the CAHSEE is a
requirement that the exam have curricular and instructional
validity:
(2) "Curricular validity" means that the examination
tests for content found in the instructional textbooks. For the
purposes of this section, any textbook or other instructional
material adopted pursuant to this code and consistent with the
state's adopted curriculum frameworks shall be deemed to satisfy
this definition.
(3) "Instructional validity" means that the examination
is consistent with what is expected to be taught. For the purposes
of this section, instruction that is consistent with the state's
adopted curriculum frameworks for the subjects tested shall be
deemed to satisfy this definition (Education Code Section 60850, f,
2 & 3).
The evidence that EL high school students do not receive the
same instruction or have access to the same range of courses as
their English-speaking peers puts in serious question the
curricular validity of these tests for English learners. Moreover,
evaluation of the early administrations of the test found that
passing rates on the math exam are significantly correlated with
completion of Algebra I (Wise, et al., 2002). Yet, EL students are
often on a high school trajectory of ELD and basic classes that
does not include algebra (Callahan, 2003). Furthermore, the
condition of these students as English learners—students who
by definition do not have the same level of understanding of all-
English instruction as fluent English proficient
students—raises questions about the instructional validity of
the exam. This is particularly true for EL students in classrooms
with teachers who do not have special certification or preparation
in English learner teaching strategies. Unfortunately, the dearth
of such prepared teachers (discussed in Section 2) is even greater
at the secondary than the elementary school level.
Accommodations
When English tests are used to assess English learners, it is
common practice in many states to use accommodations. Examples of
test accommodations include using a parallel form of the same test
content in the native language, administering the test in small
groups, repeating directions, having a person familiar with the
child’s language and culture give the test, giving more time
breaks, reading questions aloud in English, translating directions,
extending the session over multiple days, simplifying directions,
and using word lists or dictionaries (National Research Council,
1999: 218). California, however, does not allow accommodations for
those EL students who have been here for over one year. CDE
guidelines state:
English Language Learners may use nonstandard
accommodations only if the local board of education adopts a policy
before testing begins that includes the criteria each school is to
use to identify ELLs eligible to use nonstandard accommodations.
… After the policy is adopted English Language Learners who
will have been enrolled in the district less than 12 months when
testing begins may use any of the nonstandard accommodations
including having the directions translated and using bilingual
dictionaries. ELLs who have been in the district for 12 months or
more may not use nonstandard accommodations unless they have IEPs
or Section 504 Plans that include accommodations.” (Note 29) Yet since the API
index counts only those English learners who have been here for
over one year, the API does not include assessment results for
students who were allowed by a local school board to use
accommodation practices.
The need for making accommodations available by at least
allowing EL students additional time is clear from inspection of
the data on the number of items and the time allotted, according to
a table available on the CDE website (Note 30). For example, the reading vocabulary
section of the test, at each grade level, has 30 items given in an
allotment of 20 minutes and for reading comprehension there are 54
items to be completed in 50 minutes for most grades. This pattern
is also found in the mathematics items. That is to say, the speed
of the test is less than one minute per item. While this may be
sufficient for native speakers of English, this is hardly
sufficient for most English Learners, particularly given that tests
such as the SAT 9 purport to be tests of achievement (or
”power”), not of speed.
(4) Inadequate instructional time to accomplish learning
goals
There is a significant body of research that shows a clear
relationship between increased time engaged in academic tasks and
increased achievement. Carroll (1963) devised the classic model
that showed learning is a function of the amount of time needed to
learn something divided by the amount of time allotted to learn it.
Others have attempted to make sense of the different ways that time
can be used productively in learning. Berliner (1990) argues that
“time on task” is different from “academic
learning time”, with the latter resulting in greater
achievement gains than the former. Karweit (1989) has shown that
“engaged time” on task, which is akin to academic
learning time, is more important than simply the time allotted, as
in the Carroll model. All of these models, however, suggest that
there is a relationship between time and learning, and that
learning increases when students are optimally engaged in learning
activities for greater amounts of time. Notwithstanding the
importance of time for learning, there are many ways in which
English learners experience less time on academic tasks than other
students:
- With the passage of Proposition 227, English learners who
enroll in a California school for the first time must remain in a
structured English immersion program for at least 30 days before
being assigned to a permanent classroom. In a recent study of
schools implementing the proposition, many teachers complained that
they did not know what to do with students during this interim
period and that a great deal of instructional time was lost trying
to accommodate students who would not be continuing on in the same
classroom. Particularly where parents had sought a waiver to have
their child attend a bilingual classroom, teachers reported not
knowing how to instruct these students. They lacked the necessary
curricula and materials for the 30 days of all-English instruction
before they began what would be their bilingual program for the
remainder of the school year (Gándara et al, 2000).
- A common way that elementary schools organize instruction for
English learners is to take them out of their regular classes for
English language development. This strategy has been demonstrated
to create further inequities in the education of “pulled
out” students because they miss the regular classroom
instruction (Cornell, 1995; Fleishman & Hopstock, 1993;
Anstrom, 1997). Nevertheless, the practice continues to be
relatively routine for English learners. There is generally no
opportunity for students to acquire the instruction they have
missed during the pull out period (Lucas, 1997; Ovando &
Collier, 1998).
- In secondary schools English learners are often assigned to
multiple periods of English as a Second Language (ESL) classes
while other students are taking a full complement of academic
courses. Commonly, when not enough courses are available in either
SDAIE or other formats, students are given shortened day schedules,
resulting in significantly less time devoted to academic
instruction (Olsen, 1997).
- Classrooms with large numbers of English learners also have
fewer assistants in them to help the teacher provide individualized
time for the students. Table 8 shows the number and types of person
hours devoted to classrooms by percent EL.
Table 8. Hours of Assistance on Instructional Activities in
Classrooms of Teachers in Grades 1-4 by Type and Classroom
Concentration of English Learners, 1999-2000 (Mean hours)
| Percent English Learners in the classroom |
Regular aides |
Special education aides |
LEP or bilingual aides |
Parents or adults |
Students |
Other specialists |
Total |
| 0 |
3 |
2 |
<1 |
4 |
1 |
1 |
11 |
| 1-25 |
3 |
1 |
<1 |
2 |
1 |
1 |
8 |
| 26-50 |
2 |
<1 |
2 |
1 |
1 |
<1 |
7 |
| 51-100 |
3 |
<1 |
2 |
<1 |
1 |
1 |
7 |
| Total |
2 |
1 |
1 |
2 |
1 |
1 |
8 |
NOTE: Results are weighted.
SOURCE: 2000 Class Size Reduction Teacher Survey (N=774).
While the district is apt to provide more bilingual aide time
for classrooms with high percentages of English learners, there is
significantly less time spent in these classrooms by parents or
other adults. The result is that classrooms with no or few English
learners enjoy more adult time in the classroom, which means that
more of these children will receive individualized instructional
attention, exacerbating the gaps in instruction and achievement
outcomes between English learners and English speakers. While it is
not necessarily the school’s or the district’s
“fault” that some schools enjoy more parent
participation, it is a fact that must be considered in distributing
resources among schools. Furthermore, when EL students are taught
by bilingual teachers these teachers are provided with much less
paraprofessional assistance than their non-bilingual colleagues. In
the view of many teachers this constitutes “penalizing”
the most prepared teachers, and their students, for their extra
expertise.
English learners in California are also more likely to be
assigned to multi-track year round schools designed to accommodate
more students on a campus. The year round plan that accommodates
the most students is Concept 6, a schedule in which students attend
school for only 163 days per year, instead of the 180 mandated by
state law (Note 31). As
Table 9 shows, English learners comprise fully half of the students
assigned to Concept 6 schools. Students on the Concept 6 calendar
attend school for 4 months twice a year, with two month breaks in
between. This provides English learners less time to assimilate
critical academic material and to be exposed to English language
models. Just as important, however, is the loss of learning that
occurs with two months breaks in school every 4 months. A
significant body of research has now established that low income
children (and English learners) are more disadvantaged by these
lengthy breaks from school than middle income children. There is a
demonstrably negative effect on their achievement (Cooper et al.,
1996). Thus, the very students who need the most exposure to
schooling, to English language models, and to opportunities to
“catch up” to their English speaking peers are more
likely to be assigned to school calendars that provide them with
fewer school days than other students and less exposure to English
in a school setting.
Table 9. Distribution Characteristics of California Schools,
2001 Percent English Language Learner Enrollment by School
Calendar
| |
|
Calendar |
| Measure |
Statistic |
Traditional/ Single-Track |
Multi-Track Not Concept 6 |
Multi-Track Concept 6 |
| Percent English Learners |
Mean |
21 |
36 |
51 |
| |
Median |
15 |
35 |
53 |
| |
Number of Schools |
5,913 |
735 |
221 |
Source: California Department of Education,
Policy and Evaluation Division
(http://cdedata.com.hosting.pacbell.net/api2001base/dbapi01b.zip)
and School Facilities Planning Division
(http://www.cde.ca.gov/facilities/yearround/direct00.htm)
(5) Inequitable access to instructional materials and
curriculum
All students need appropriate instructional materials. While
some might argue that textbooks and other appropriate learning
materials are not essential to learning, Oakes and Saunders (2002)
have argued cogently that the preponderance of research evidence
demonstrates a clear link between appropriate materials and
curriculum and student academic outcome. English learners, however,
need additional instructional materials in two areas. First, all
English learners need developmentally appropriate materials to
learn English and to master English Language Development standards.
Second, English learners receiving primary language instruction
need appropriate materials in their native language. However, the
evidence suggests that many are not gaining access to such
materials. In the second year report of the AIR study, researchers
report that 75% of the teachers surveyed said they “use the
same textbooks for my English learner and English only
students” and fewer than half (46%) reported using any
supplementary materials for EL students (Parrish, et al., 2002, p.
IV-34). This raises the question of how much EL students can be
expected to learn without materials adapted to their linguistic
needs. It is not particularly surprising then that only 40.9% of
teachers report they are “able to cover as much material with
EL students as with EO students” (Parrish, et al., 2002, p.
IV-35). There is ample evidence in the research literature that
when students cover less material than their peers, their skills
decline relative to other students and they are prone to be placed
in low academic groupings or tracks where educational opportunities
are limited (Barr & Dreeben, 1983; Oakes, 1985; Goodlad, 1984;
Gamoran, 1992).
The quality of instructional materials appears to differ by
concentration of English learners in the school as well. Data from
the Harris survey show that teachers with high percentages of
English learners are less likely than teachers with low percentages
of English learners to have access to textbooks and instructional
materials, in general, and materials needed by English learners in
particular. Almost half of teachers with high percentages of
English learners report that the textbooks and instructional
materials at their schools were only fair or poor compared to 29
percent of teachers with low percentages of English learners (Table
10). Teachers with high percentages of English learners were also
almost twice as likely as teachers with low percentages of English
learners to report that the availability of computers and other
technology was only fair or poor. Moreover, almost two-thirds of
teachers with high percentages of ELs in their classes reported not
enough or no reading materials in the home language of their
children and more than one quarter reported that they did not have
any or enough reading materials at students reading levels in
English.
Table 10. Condition and Availability of Instructional
Materials in California Schools by Percentage of English Learners
in Teachers’ Schools or Classrooms, January 2002 (percent of
teachers reporting condition)
| |
School EL |
|
| |
25% or less |
Over 25% |
Total |
| Reported by all teachers (N=1071) |
|
|
|
| Textbooks and instructional materials are ONLY
FAIR OR POOR |
14 |
22 |
17 |
| Availability of computers and other technology
is ONLY FAIR OR POOR |
26 |
40 |
31 |
| |
Classroom EL |
|
| |
30% or less |
Over 30% |
|
| Reported by teachers who have EL students in
their classes (N=829) |
|
|
|
| Not enough or no reading materials in home
language of children |
44 |
68 |
51 |
| Not enough or no reading materials at students
reading levels in English |
19 |
29 |
22 |
NOTE: Results exclude respondents who did not
answer question or answered “not sure.” Results are
weighted. All column differences are statistically significant at
.05 level or better.
Source: Harris Survey of a Cross-Section of California School
Teachers, January 2002.
Weak Curriculum
There is a significant body of research on tracking that
connects the rigor of instruction to students’ academic
outcomes (Mickelson, 1999; Oakes, 1985; Page; 1990). The negative
effects of low-track instruction are doubly detrimental for ELs who
enter the classroom already at a linguistic disadvantage (Callahan,
2003; Harklau, 1994; Katz, 1999). There is a common perception that
English learners are clustered in the early years of school, and so
most attention is applied to students in this age group. However,
about one-third of English learners in California are found in
grades 7 – 12. And, these students are often shortchanged by
their schools because of lack of appropriate coursework offerings
or materials to support courses for English learners. In secondary
schools, English learners are often assigned to multiple periods of
ESL or ELD classes while other students are taking a full
complement of academic courses. Commonly, when not enough courses
are available in either SDAIE or other formats, students are given
shortened day schedules, resulting in significantly less time
devoted to academic instruction (Olsen, 1997).
We selected a random sample of transcripts of secondary English
learners from two different northern California districts. In
district #1, we compared a random sample of English learners with a
random sample of English speaking students. For English only
students (20) with GPAs from 1.6 to 4.1, 58% of their courses were
college preparatory. For the English learners (8), with GPAs from
1.3 to 3.3 (this was the upper bound of the GPA range for English
learners), 21% of their courses were college preparatory. The
following are samples of English learner programs for the sophomore
and senior years:
District #1
Saul (2 years in U.S., attended 9th grade in Mexico where he was
in a college preparatory curriculum and took advanced mathematics
courses) Sophomore year (2001):
Period 1: No class
Period 2: Language Development 1
Period 3: Language Development 2
Period 4: Native Spanish1
Period 5: U.S. History (in Spanish)
Period 6: Math A (general, low level)
Period 7: Weightlifting
(two courses meet college preparatory requirements: Spanish and
U.S. History. No science is provided.)
Jose Luis (1 year in the U.S. Uneven academic history prior to
immigration) Sophomore (2001):
Period 1: No class
Period 2: Language Development 1
Period 3: Language Development 1
Period 4: General Math (in English)
Period 5: Native Spanish 1
Period 6: Drawing 1
Period 7: No class
(One class prepares student for college requirements: Spanish.
No science or social science offered. Student failed English only
math because he could not understand the teacher.)
District #2
Marcos (Long term EL student, enrolled in California schools
prior to entering high school). Sophomore (2000):
Period 1: English 10 SDAIE
Period 2: World History SDAIE
Period 3: Pre Algebra A SDAIE
Period 4: Court Sports
Period 5: Integrated Science 2 SDAIE
Period 6: ELD 5
(Only two courses could be used to meet college preparatory
requirements: World History and Integrated Science as an elective,
not as a science course. Student never took a college preparatory
science, math or English course through the junior year of high
school.)
Marisela(Long term EL student, enrolled in California schools
prior to entering high school) Senior year (2002):
Period 1: Power English
Period 2: Weight training
Period 3: ELD 5C
Period 4: Business Math
Period 5: Consumer Foods
Period 6: Floral Design
(None of the student’s courses meets college preparatory
criteria. The student took no laboratory science or math beyond
algebra. 1 which she failed and received no credit.)
These are students who have been attending California schools
with caring administrators and school personnel, but the schools
did not have the resources –human or otherwise—to
provide an appropriate program of study for these students. They
were selected randomly from among a pool of students like them for
illustrative purposes, but they represent typical scenarios in many
of California’s high schools.
Because the state does not effectively monitor the quality of
instruction that English learners receive, or the amount of time
they spend in Structured English Immersion settings, we do not know
to what extent the educational services provided for these students
meet high standards of quality. We can guess at this figure, given
the large numbers of unprepared teachers who teach them. It is
worth noting, however, that more than 82,000 English learners in
California receive no special instruction whatsoever. For
some of these students this is based on parental request, but even
this requires greater scrutiny. The AIR study of the implementation
of Proposition 227 (Parrish, et al, 2002) noted that there remained
a great deal of confusion among parents about what options existed
for them, and that “in some cases, teachers are discouraged
from discussing educational alternatives for students” (p.
IV-41). In this environment, some parents are certainly making
uninformed decisions about their children’s educational
program. The state has not monitored the extent to which schools
and districts provide full disclosure to parents about the programs
they may and do offer.
Over-placement in Special Education resulting in weak
curriculum
The persistent and pervasive inequities in access to well-
prepared teachers, school resources and facilities, appropriate
assessment and time to accomplish learning goals result in large
and growing gaps in achievement for English learners vis-à-vis
their English speaking peers, and ultimately to misplacement into
some special education classes. In the consent decree resulting
from the Diana v California State Board of Education
(U. S. D. C., ND, Cal.1970), a class action suit on behalf of
English learners inappropriately placed in special
education, the state agreed to the following:
- To test Mexican American children in their own language and in
English
- To test them on the non-verbal sections of intelligence
tests
- To re-test all Mexican American who are in Educable Mentally
Retarded (EMR) classes using non-verbal sections of intelligence
tests
- Develop and norm a test of IQ that reflects Mexican American
culture
- Require school districts throughout the state that show a
significant disparity between their overall district racial-ethnic
representation and the racial-ethnic representation in their EMR
classes to submit an explanation for the overrepresentation.
Thirty years hence the State of California has still not acted
to implement the consent decree with respect to the development of
appropriate assessment for English learners that could stem the
over diagnosis and placement of these students in special
education. Nor does California keep reliable data on the numbers of
EL students in special education. About to be published is a study
based on data from eleven school districts and over 700,000
students in the Los Angeles area for the 1998-99 school year. The
researchers, Artiles and Rueda (in press) report that “ELs
are over-represented in special education, particularly in specific
learning disabilities (SLD) and language and speech impairment
classes (SLI), especially at the secondary grade level where
language support is minimal” (pg.2). Even more distressing is
that, “highly vulnerable ELs (those who have low proficiency
in both English and their primary language) are 1.5 times more
likely to be diagnosed as Speech Impaired and Learning Disabled
than their English speaking peers during the elementary school
years. During the high school years, “highly vulnerable
ELs” are twice as likely to be diagnosed as Mentally
Retarded, Speech Impaired, and Learning Disabled. The state of
being highly vulnerable –or having low proficiency in two
languages—is often a product of inadequate instruction, just
as proficiency in at least one language is the usual outcome of
schooling and this is true for all children, regardless of their
ability level. We know, for example, that many mentally disabled
children acquire a reasonable proficiency in their primary language
(Rueda, R. & Smith,1983; Whitaker, Rueda, & Prieto,1985).
Table 11 shows that English learners and highly vulnerable English
learners are significantly over-represented in special education
programs in the sampled districts.
Table 11. Percent Students in Special Education, Elementary
(K-5) & Secondary (6-12) Compared to Percent of Total School
Population by language status and White (non EL), 11 Los Angeles
Area School Districts
| |
White |
Typical EL |
Highly Vulnerable EL |
| Elementary |
(9%)* |
(28%)* |
(22%)* |
| SLI |
14% |
17% |
48%** |
| SLD |
13% |
29% |
48%** |
| |
|
|
|
| Secondary |
(12%)* |
(12%)* |
(13%)* |
| MMR |
14% |
9% |
26%** |
| SLI |
11% |
10% |
27%** |
| SLD |
9% |
17%** |
23%** |
*p<.05; **p<.01
Source: Artiles & Rueda, in press
As was the case with the 1982 report by the National Academy of
Sciences (Heller, Holtzman & Messick, 1982), an important
finding is underscored by Artiles and Rueda: where there are few if
any primary language support services offered, special education
misdiagnosis and misplacement tends to occur. This is almost
certainly related as well to the inequitable distribution of
psychologists in the schools who can meet the assessment needs of
English learners. The National Association of School Psychologists
reports that only 160 out of all school psychologists in California
report having bilingual competency. There are currently 1,949
school psychologists employed in California schools. If all of the
bilingual psychologists were employed in the schools (which they
almost certainly are not) then only 8% of psychologists would be
bilingual and capable of conducting an assessment in a
student’s primary language. And, if all of these
psychologists were assigned only to English learners, then 8% of
the psychologists would be assessing 25% of the students.
Placement in special education, especially when it is not
warranted, can have devastating effects on students’ access
to opportunities later in life. Evidence has existed for years
documenting the massive rates of high school non-completion,
underemployment, poverty, and adult marginalization of special
education students after they leave high school (Guy, Hasazi, &
Johnson, 1999). Placed in a special education track, it is unlikely
for students to rejoin the mainstream. Robert Peckham, the
presiding judge for the Diana case, summarized the evidence
on the effectiveness of California’s special education
program, calling it a “dead-end educational program”
(Crawford v. Honig, 1988).
(6) Inequitable access to adequate facilities
While it has been notably difficult to establish a firm link
between the quality and condition of school facilities and the
educational outcomes for students-- largely because the quality of
school facilities is so highly correlated with wealth of the
students and communities that schools serve – there is
considerable consensus that it is difficult to both teach and learn
in grossly inadequate facilities (Ortiz, 2002). Equally important,
as we have already established, the conditions of schools are also
related to teacher turnover. Teachers do not want to teach in
dirty, dangerous, and uncomfortable conditions, and so they leave
when they can. And, given that it is exceptionally difficult, if
not impossible, to effect school reform without a stable base of
teachers, school facilities certainly play an important, albeit
indirect, role in student achievement.
Teachers of English learners are more apt than teachers of
English speakers to respond that they do not have facilities that
are conducive to teaching and learning. In the Harris survey close
to half of teachers in schools with higher percentages of English
learners reported the physical facilities at their schools were
only fair or poor, compared to 26 percent of teachers in schools
with low percentages of English learners (Table 12). Teachers in
schools with high percentages of English learners were 50 percent
more likely to report bathrooms that were not clean and open
throughout the day and having seen evidence of cockroaches, rats,
or mice.
Table 12. Condition of Facilities of California Schools by
Percentage of English Learners in Teachers’ Schools, January
2002 (percent of teachers reporting condition)
| |
25% or less |
Over 25% |
Total |
| The adequacy of the physical facilities is ONLY
FAIR OR POOR |
26 |
43 |
32 |
| Bathrooms ARE NOT clean and open for throughout
day. |
13 |
23 |
17 |
| HAVE seen evidence of cockroaches, rats, or
mice in past year. |
24 |
34 |
28 |
Note: Results exclude respondents who did not
answer question or answered “not sure.” Results are
weighted. All column differences are statistically significant at
.05 level or better.
Source: Harris Survey of a Cross-Section of California School
Teachers, January 2002 (N=1071).
ECLS data show the same picture with regard to facilities. More
than a third of principals in schools with higher concentrations of
English learners reported that their classrooms were never or often
not adequate, compared to 8 percent of principals with low
concentration of EL students (Table 13). (Note 32)
Table 13. Characteristics of California Elementary School
Facilities by EL Concentration, Spring1999
| |
25% or less |
More than 25% |
Total |
| Principal questionnaire responses: |
|
|
|
| Classrooms never or often not adequate (Note 33) |
8 |
35 |
19 |
Note: Results are weighted (S2SAQW0).
Source: ECLS base year data for California public elementary
schools (N=69).
Such conditions not only make it more difficult to teacher
English learners, they also make it difficult to retain teachers
since, as we showed earlier, a considerable body of research finds
that teachers are more likely to leave schools with poor working
conditions.
(7) Intense segregation into schools and classrooms that
place them at particularly high risk for educational
failure
Research on desegregation has established that minority students
who are schooled in desegregated settings tend to have better
occupational outcomes and overall life chances (Wells & Crain,
1994; Crain & Strauss, 1985). Sociologists often explain this
phenomenon as the impact of social capital –access to
important social networks—on student outcomes (see for
example, Stanton-Salazar, 1997). Therefore, one reason to be
concerned about racial, ethnic, or linguistic isolation is the
effect it has on limiting access to important social networks.
However, a more immediate impact of linguistic isolation is the
lack of appropriate English language models, which can result in
both reduced opportunities to hear and interact with the language,
and fewer opportunities to understand the ways in which the
language is actually used in social and academic contexts (August
& Hakuta, 1997). Both are important features in the development
of high levels of linguistic skill.
We argue that the concentration of English learners in
classrooms and schools in California compromises their opportunity
to receive an education that is comparable in quality and scope to
that of their non-EL peers because: (1) the lack of peer English
language models limits the development of English; (2) the lack of
models of children who are achieving at high or even moderate
levels inhibits academic achievement, (3) the inequitable
environmental conditions and resources of segregated classrooms and
schools, and (4) the lack of highly qualified, experienced,
teachers in these particular classrooms depress learning.
The first source of inequity stems directly from the segregation
itself --English learners are more likely to attend classes and
schools surrounded by other students who are not proficient in
English. This hurts English learners’ ability to become
proficient in English because research has shown that the
composition (relative numbers of English-language learners and
fluent English speakers) and structure (opportunities for
interaction) of the classroom can inhibit meaningful second
language acquisition (Hornberger, 1990; Wong Fillmore, 1991).
Moreover, a recent California study found that the higher
concentrations of English learners in schools, the lower rates of
reading development in first grade (Rumberger & Arellano,
2003).
The educational achievement of English learners is also hurt by
their segregation because they are less likely than other students
to be surrounded by peers who excel in school. As shown in Table 3,
classrooms with high concentrations of English learners also have a
higher number of students who are below grade level in reading and
math than classrooms with low concentrations of English learners.
Research has shown that the academic achievement of peers
influences students’ own academic achievement, in part,
because students learn from each other (Epstein & Karweit,
1983; Hanushek, Kain,Markman, & Rivkin,2001; Hoxby, 2001;
Mounts & Steinberg, 1995; Hurd, in press). Thus, the
concentration of English learners in California’s schools and
classrooms not only makes it more difficult for them to learn
English, it also makes it more difficult for them to achieve
academically.
Lack of integration with native English speakers, both at the
school-site and the classroom level can therefore prove problematic
for ELs as well. English learners are highly segregated among
California’s schools, and thus isolated from the language
models and social capital necessary to ensure success in school
(Harklau, 1994; Olsen, 1997). While most schools have some English
learners, the vast majority of these students attend a relatively
small percentage of public schools. Thus, English learners are much
more likely than their English-only peers to attend schools with
large concentrations of EL students. As shown in Table 14, while
twenty-five (25) percent of all students in California attend
elementary schools in which a majority of the students are English
learners, more than half of all English learners (55%) are enrolled
in such schools. At the middle school level, only 8 percent of the
schools have more than 50 percent of the English learners. Very few
high schools have such high concentrations of English learners.
Nonetheless, almost half of all EL students attend high schools
with more than 25 percent English learners. Thus, the distribution
of English learners across schools is uneven and these students
tend to be clustered in a relatively small percentage of
schools.
Table 14. Schools, Students, and English Learners by
Concentration of English Learners and School Level, 1999-2000
(Percent Distribution)
| Percent English Learners |
Elementary |
Middle |
High |
| Schools |
Students |
ELs |
Schools |
Students |
ELs |
Schools |
Students |
ELs |
| 0 |
6 |
1 |
0 |
<1 |
1 |
0 |
8 |
2 |
0 |
| 1-25 |
51 |
48 |
15 |
65 |
62 |
30 |
73 |
76 |
49 |
| 26-50 |
24 |
26 |
30 |
26 |
28 |
44 |
17 |
21 |
46 |
| 51-100 |
19 |
25 |
55 |
8 |
9 |
25 |
1 |
1 |
5 |
| Total percent |
100 |
100 |
100 |
100 |
100 |
100 |
100 |
100 |
100 |
| Total number |
5,306 |
3,124,107 |
979,854 |
1,158 |
1,059,767 |
232,481 |
909 |
1,538,617 |
237,129 |
Note: ELs = English learners.
Source: CBEDS and Language Census.
English learners are even more concentrated at the classroom
level. In 2000, researchers from the California Class Size
Reduction Study surveyed a representative sample of California
teachers who taught grades 1-4 (Stecher & Bohrnstedt, 2002).
They found that more than three quarters of all teachers had at
least some English learners in their classrooms, and almost one-
quarter taught in classrooms with more than 50 percent English
learners (Table 15). Even more striking, almost two-thirds of
English learners enrolled in grades 1-4 attended classrooms in
which more than 50 percent of their fellow students were English
learners. Thus, while classrooms in grades 1-4 enrolled an average
of 6 English learners (see Table 16) in 2000, the distribution of
these students across classes was highly uneven.
Table 15. Teachers, Students, and English Learners in Grades
1-4 by Classroom Concentration of English Learners, 2000 (Percent
Distribution)
| Percent English Learners in the classroom |
Teachers |
Students |
English Learners |
| 0 |
24 |
23 |
0 |
| 1-25 |
38 |
39 |
17 |
| 26-50 |
16 |
16 |
20 |
| 51-100 |
22 |
22 |
63 |
| Total percent |
100 |
100 |
100 |
Note: Results are weighted.
Source: 2000 Class Size Reduction Teacher Survey (N=774).
If students were clustered into these classrooms in order to
provide core academic instruction in the primary language and
mainstreamed for part of the day to receive instruction in English
(preferably in highly interactive and non-high stakes settings like
arts, music, physical education), the segregation of EL students
would not only be defensible, but would constitute a valid
educational treatment. However, in the wake of Proposition 227,
most English learners are simply segregated into classrooms
populated disproportionately by other English learners where the
opportunity to learn both English and academic content is
compromised by the lack of appropriate models and ins |
